pincer-workflow 0.2.2 → 0.3.0

package/template/.claude/commands/pincer-code.md

@@ -5,8 +5,9 @@ argument-hint: "Ticket numbers (optional — defaults to all open tickets in ord
 
 # /pincer-code — Ticket Implementation
 
-You are implementing the tickets in `tickets/` sequentially. Mostly autonomous: after the
-user confirms the starting point, run continuously and report progress between tickets.
+You are implementing the tickets in `tickets/` sequentially. The approved PRD, ticket
+breakdown, and existing session authorization define the work; run continuously and report
+progress between tickets unless a material scope or design decision appears.
 
 **Initial request:** $ARGUMENTS
 
@@ -14,21 +15,22 @@ Ticket state lives in the ticket file's frontmatter and is written **only** by
 `scripts/pincer-ticket.sh` (`start` → `verify` → `done`). `verify` runs the ticket's
 Verification block and stamps a receipt only on a green exit; `done` refuses without a
 receipt that matches the current check, or with unticked acceptance criteria. Never edit
-`status`, `started`, `verified`, or `finished` by hand — on Claude Code a hook blocks it.
+`status`, `started`, `last_check`, `verified`, or `finished` by hand — on Claude Code a hook blocks it.
 
 ## Before the loop
 
 Run `scripts/pincer-status.sh`. It lists every ticket's state, what is blocked, elapsed
 build time from the clock, and the next action. If a ticket is `in_progress`, you are
 resuming: read it, check `git status` / `git diff` for uncommitted work, and continue
-from wherever the receipt says you are. Confirm the starting point with the user, then go.
+from wherever the receipt says you are. Do not ask the user to reconfirm unchanged,
+previously authorized work.
 
 ## Loop (per ticket, in dependency order)
 
 1. **Start:** `scripts/pincer-ticket.sh start T-{NN}` — refuses while a `depends_on` ticket
    isn't done, and stamps the start time. Read the ticket and the files it references.
    Announce: "Starting T-{NN}: {title}."
-2. **Implement.** Follow the conventions in `CLAUDE.md` and the PRD's architecture and
+2. **Implement.** Follow the conventions in `AGENTS.md` and the PRD's architecture and
    visual direction. Installing a dependency not named in the PRD's architecture is a
    stop-and-ask: verify it's the real package on the registry (linked repo, downloads —
    hallucinated names get typosquatted), say why it earns its place, and wait for a yes.
@@ -41,26 +43,27 @@ from wherever the receipt says you are. Confirm the starting point with the user
 4. **Self-review the diff** before committing: silent failures (empty catches,
    un-awaited promises), leftover debug code, drift from the ticket's acceptance criteria.
    Then a security sweep of the same diff:
-   - No secret values: run
-     `git diff | grep -iE '(api[_-]?key|secret|token|password)[[:space:]]*[:=]'`
-     and treat any hit that isn't a `process.env` reference or a name in
-     `.env.example` as a blocker.
+   - Check for secret-like assignments without printing values. If a scanner reports a
+     possible secret, report only its file and line until the value is safely redacted;
+     environment references and names in `.env.example` are allowed.
    - External input touched by this diff is validated server-side, and untrusted
      content (user input, LLM output) is escaped where rendered — per the
-     Security defaults in `CLAUDE.md`.
+     Security defaults in `AGENTS.md`.
    - No error path leaks internals (stack traces, key names with values) to the client.
    If the review changed code, run `verify` again — the receipt must match the code you commit.
 5. **Close the ticket:** tick every verified acceptance-criteria checkbox (`- [ ]` → `- [x]`;
    editing the checkboxes is allowed), then `scripts/pincer-ticket.sh done T-{NN}`. A
    criterion that was cut is a scope change to record in the PRD, not a box to skip.
-   Commit code and ticket file together: `git add -A && git commit -m "T-{NN}: {title}"`.
+   Inspect `git status --short`, preserve pre-existing staged work, and stage only the
+   explicit paths changed for this ticket plus its ticket file. Review `git diff --cached`
+   before committing as `T-{NN}: {title}`.
 6. Give a one-line progress update using the elapsed figure from
-   `scripts/pincer-status.sh` ("T-02 done, 3 remaining, 38m elapsed of 75m") and continue.
+   `scripts/pincer-status.sh` ("T-02 done, 3 remaining, 38m elapsed") and continue.
 
-## Timebox rules
+## Budget rules
 
-- The budget is ~75 minutes of build time, measured by `scripts/pincer-status.sh` from
-  the first ticket's start stamp — never estimated. If the remaining tickets won't fit,
+- If the user set `PINCER_BUILD_BUDGET_MIN` or stated another budget, use the elapsed
+  figure from `scripts/pincer-status.sh` rather than estimating. If the remaining tickets won't fit,
   stop and propose a scope cut: which remaining tickets to drop or shrink. Cutting scope
   deliberately beats an unfinished mess — record the cut in the PRD's Out of Scope.
 - If a ticket reveals the plan was wrong, stop and say so rather than silently diverging.

package/template/.claude/commands/pincer-evaluate.md

@@ -10,10 +10,15 @@ run the pipeline, then present results.
 
 ## Steps
 
-1. Run `scripts/pincer-status.sh`. Every ticket should be `done` with a receipt; if one
+1. Run `scripts/pincer-status.sh`. Review only tickets associated with the selected
+   PRD. Every such ticket should be `done` with a current receipt; if one
    is still open or in progress, stop and ask whether it was cut (then it goes in the
    PRD's Out of Scope) or should be finished first via `/pincer-code`. Then get the full
-   diff of the session: `git log --oneline` and `git diff <first-commit>..HEAD`.
+   diff of the change: identify the actual base commit before this change from
+   its ticket commits and recorded context. If it cannot be established, resolve
+   that uncertainty before claiming a complete review. Record full commit IDs for
+   `base` and `candidate` (`git rev-parse HEAD`), then review `git diff <base>..<candidate>`.
+   Require a clean candidate before review, excluding only the notes being written.
 2. Dispatch a `code-quality-reviewer` agent with: the diff, the PRD's Success Criteria and
    Scope sections, and the list of tickets. If the diff is large, split by area and
    dispatch two in parallel. (No subagents on this platform? Review the diff yourself
@@ -23,10 +28,11 @@ run the pipeline, then present results.
 4. If the project has a UI, look at it — don't only read the code. Start it, open it in
    the browser (screenshot via Chrome DevTools MCP if available), and check it against
    the PRD's Visual Direction and Success Criteria. Note anything visibly broken or off.
-5. Run a mechanical security audit (cheap, ~2 min — do all of these):
-   - Whole history, not just the tree:
-     `git log -p | grep -iE '(api[_-]?key|secret|token|password)[[:space:]]*[:=]'` —
-     a secret committed then deleted is still leaked.
+5. Run a mechanical security audit:
+   - Inspect the relevant history with a secret scanner that redacts values, when one is
+     available. Otherwise review likely locations without copying candidate values into
+     output. Report file, line, and remediation only; a secret committed then deleted is
+     still leaked.
    - `.gitignore` covers `.env*` (except `.env.example`), and `git ls-files | grep -i env`
      shows only `.env.example`.
    - `npm audit --omit=dev` (or the ecosystem's equivalent) — report high/critical only.
@@ -37,9 +43,22 @@ run the pipeline, then present results.
 7. Present findings as a short list with `file:line` references, ordered by severity.
    Security findings always rank above style-adjacent ones. For each, say whether you
    recommend fixing now (within the timebox) or noting as known-issue.
-8. Fix what the user approves (or everything clearly broken, if time allows), verify,
-   and commit as `review: fixes`.
-9. Close out: write a brief `NOTES.md` at the repo root — what was built, what was cut
+8. Fix findings clearly within the authorized PRD through a new ticket associated with
+   that PRD. Use `pincer-ticket.sh` to start, verify, and close it, then make a scoped
+   `T-{NN}: {title}` commit. Ask only when a fix changes scope, architecture, or another
+   material decision; never make an ad-hoc `review: fixes` commit.
+9. Close out: write a brief `NOTES.md` at the repo root with frontmatter:
+   ```yaml
+   ---
+   prd: .prd/prd-vN.md
+   base: <full reviewed base commit ID>
+   candidate: <full reviewed candidate commit ID>
+   ---
+   ```
+   Record the candidate before the separate NOTES commit. Status accepts a later
+   commit only when its diff from the candidate changes solely `NOTES.md`; changes
+   to source, tickets, or PRD require reevaluation. Legacy notes without these
+   references do not establish readiness. Then describe what was built, what was cut
    and why, known issues, and what you'd do next with more time. Then a **Handover**
    section, written for the stranger who inherits this repo in six months: how to get
    oriented (which file to read first), what each dependency is for and why it earned

package/template/.claude/commands/pincer-narrow.md

@@ -5,22 +5,26 @@ argument-hint: "Path to PRD (optional — defaults to the latest in .prd/)"
 
 # /pincer-narrow — PRD to Local Tickets
 
-You are decomposing the PRD into small, independently verifiable tickets stored as local
-markdown files (no external tracker needed). Target: 4–7 tickets that fit a ~75-minute
-build window.
+You are decomposing the PRD into coherent, independently verifiable tickets stored as
+local markdown files (no external tracker needed). Ticket count and size follow the
+change's dependencies and risk, plus any budget the user supplied.
 
 **Initial request:** $ARGUMENTS
 
 ## Steps
 
-1. Run `scripts/pincer-status.sh`. If tickets already exist, ask before adding to them —
-   new tickets continue the numbering, existing ones are never renumbered. Then read the
+1. Run `scripts/pincer-status.sh`. If tickets already exist for another PRD, leave them
+   as history. New tickets continue the numbering and existing ones are never renumbered.
+   If tickets already exist for this PRD, extend them only when the current request already
+   authorizes that work; otherwise present the concrete addition before asking. Then read the
    PRD (`$ARGUMENTS` or the latest `.prd/prd-v*.md`). If its status isn't `draft`, ask
    which PRD to use.
 2. Decompose into tickets. Rules:
-   - Each ticket is one coherent unit: sized S or M, never L. Split anything larger.
-   - Ticket 1 is always the walking skeleton: project scaffold + a thin end-to-end slice
-     that runs. Everything after builds on a working base.
+   - Each ticket is one coherent unit. Use S, M, or L as relative scope indicators and
+     split work when that improves dependency order, verification, or ownership.
+   - For greenfield work, use a walking skeleton when it reduces integration risk. For
+     brownfield work, begin with the smallest protected vertical change; add a
+     characterization ticket before changing load-bearing code that lacks coverage.
    - Order by dependency; note blockers explicitly ("depends on T-01").
    - Every ticket gets a runnable command in its Verification block — a fenced `bash`
      block that exits 0 only when the ticket is done. `scripts/pincer-ticket.sh verify`
@@ -32,20 +36,22 @@ build window.
      LLM output) gets an acceptance criterion for the reject path — what invalid
      input produces (e.g. "empty goal → 400 with a clear message"), not only the
      happy path.
-   - The walking skeleton (T-01) includes `.gitignore` covering `.env*` (except
-     `.env.example`) and an `.env.example` naming any required secrets — before
-     any secret can exist in the repo.
-   - Brownfield: a ticket that modifies load-bearing code with no test coverage
-     is preceded by a characterization ticket — a test that pins the current
-     behavior before any ticket is allowed to change it.
+   - A greenfield setup ticket includes `.gitignore` covering `.env*` (except
+     `.env.example`) and an `.env.example` naming any required secrets before any
+     secret can exist in the repo. In brownfield repositories, preserve and verify
+     the existing ignore and environment conventions.
 3. Write each ticket to `tickets/T-{NN}-{slug}.md` using
-   `.claude/references/ticket-template.md`, with `status: open`. The other state fields
-   (`started`, `verified`, `finished`) are added later by `scripts/pincer-ticket.sh` —
+   `.claude/references/ticket-template.md`, with `status: open` and an explicit
+   `prd: .prd/prd-vN.md` naming the selected PRD. Never infer this association from
+   numbering or old notes. The other state fields
+   (`started`, `last_check`, `verified`, `finished`) are added later by `scripts/pincer-ticket.sh` —
    never write them yourself.
 4. Present the ticket list (number, title, size, dependencies) as a table.
 
-**Gate (medium):** Ask for approval of the breakdown and build order. Adjust if pushed back.
+Present the concrete breakdown and build order. Reuse existing authorization for the same
+scope and order; ask only when the breakdown introduces a material decision or scope change.
 
-5. After approval, update the PRD frontmatter to `status: ticketed`, commit the tickets
-   (`git add .prd tickets && git commit`), and finish with:
+5. Once authorized, update the selected PRD frontmatter to `status: ticketed`. Inspect
+   existing staged changes, stage that PRD and the explicit new ticket paths, review
+   `git diff --cached`, and commit only those paths. Finish with:
    "Tickets ready in `tickets/`. Run `/pincer-code` to start implementing."

package/template/.claude/commands/pincer-plan.md

@@ -5,17 +5,19 @@ argument-hint: "Brief description of the task/feature (paste the full brief if y
 
 # /pincer-plan — PRD Creation
 
-You are turning a task brief into a compact PRD. This runs inside a short delivery
-timebox (~2 hours total), so discovery is brief and the PRD is lean. The PRD feeds
+You are turning a task brief into a reviewable PRD. Scale discovery and detail to
+the change's uncertainty, risk, and any time budget the user supplied. The PRD feeds
 `/pincer-narrow` next.
 
 **Initial request:** $ARGUMENTS
 
-First run `scripts/pincer-status.sh`. If a PRD already exists, say so and ask whether
-this is a new version of it (`.prd/prd-v{N+1}.md` — old versions are never overwritten)
-or a fresh start; if tickets are in progress, stop and point at `/pincer-code` instead.
+First run `scripts/pincer-status.sh`. If a PRD already exists, preserve it and select
+the next unused numeric version for this change. Use the brief and repository state to
+distinguish a revision from a new change; ask only if that distinction changes scope or
+architecture. If tickets for the current PRD are in progress, resume `/pincer-code`
+unless the user explicitly authorized a separate change.
 
-## Phase 1: Discovery (~5 min)
+## Phase 1: Discovery
 
 1. If `$ARGUMENTS` contains the brief, extract what you can before asking anything.
    Never ask a question the brief already answers.
@@ -27,9 +29,10 @@ or a fresh start; if tickets are in progress, stop and point at `/pincer-code` i
 3. If the project has a frontend, ask one design question: "What should this feel like,
    and what should it NOT look like?" Capture the answer for the Visual Direction section.
 
-Summarize your understanding in 3–5 sentences and confirm before moving on.
+Summarize your understanding in 3–5 sentences. Existing authorization in the request or
+session carries forward; ask only about an unresolved choice that materially changes the result.
 
-## Phase 2: Codebase scan (conditional, ~5 min)
+## Phase 2: Codebase scan (conditional)
 
 If the repo already contains source code, launch 1–2 `codebase-explorer` agents in parallel
 (one for architecture/structure, one for patterns relevant to the feature). Read the 2–3 most
@@ -44,7 +47,7 @@ touches are load-bearing, what test coverage protects them (run the suite, don't
 and the blast radius + rollback story for the change. Record these in the PRD's
 Architecture section. Greenfield speed assumptions do not transfer to brownfield work.
 
-## Phase 3: Architecture (~5 min)
+## Phase 3: Architecture
 
 Propose the architecture: components, data flow, integration points, and key decisions.
 - Recommend one approach; mention an alternative only when the trade-off is real.
@@ -57,23 +60,28 @@ Propose the architecture: components, data flow, integration points, and key dec
 - Verify the contract of any external API the plan builds on (one live request or the
   current official docs) before designing around it — endpoint shapes remembered from
   training data are guesses.
-- Bias every decision toward "finishable in the remaining time". Cut before you gold-plate.
+- Respect any explicit delivery budget. Record deliberate cuts in Out of Scope.
 
-**Gate (heavy):** Ask for explicit approval of the architecture before writing the PRD.
+Prepare the full draft before seeking any approval still required. The user should review a
+concrete scope and architecture; do not repeat an approval already given for the same decision.
 
-## Phase 4: Write the PRD (~5 min)
+## Phase 4: Write the PRD
 
 1. Load `.claude/references/prd-template.md` and write all core sections.
-2. Include optional sections only when they earn their space in the timebox.
-3. Save to `.prd/prd-v1.md` (create `.prd/` if needed) with frontmatter:
+2. Include optional sections when risk or the product context warrants them.
+3. Save to the next unused `.prd/prd-v{N}.md` (create `.prd/` if needed), with `N`
+   matching the filename and frontmatter:
    ```yaml
    ---
-   version: 1
+   version: {N}
    status: draft
    date: {today}
    ---
    ```
-4. If `.git/` doesn't exist, run `git init` and make an initial commit containing the
-   PRD and this `.claude/` setup — planning should be visible in the history.
+4. If `.git/` doesn't exist, run `git init`. Commit the PRD and only the intended setup
+   paths after inspecting existing staged work; planning should be visible in history
+   without absorbing unrelated brownfield changes.
 
-Finish with: "PRD saved to `.prd/prd-v1.md`. Run `/pincer-narrow` to break it into work items."
+Present the saved draft and obtain approval only when the same scope/architecture was not
+already authorized. Finish with: "PRD saved to `.prd/prd-v{N}.md`. Run `/pincer-narrow`
+to break it into work items."

package/template/.claude/commands/pincer-release.md

@@ -5,24 +5,29 @@ argument-hint: "Stage to check (optional: plan | narrow | code | evaluate — de
 
 # /pincer-release — Workflow Audit
 
-You are auditing the current repo state against `docs/dry-run-checklist.md`. Read-only:
+You are auditing the current repo state against `docs/release-checklist.md`. Read-only:
 report pass/fail, never fix anything — fixes belong to the stage commands.
 
 **Requested stage:** $ARGUMENTS
 
 ## Steps
 
-1. Read `docs/dry-run-checklist.md` — it is the source of truth for what to check.
-2. Run `scripts/pincer-status.sh` to determine which stages have run (it reads `.prd/`,
-   `tickets/`, `NOTES.md`; add `git log`). If `$ARGUMENTS` names a stage, check only up
+1. Read `docs/release-checklist.md` — it is the source of truth for product-candidate
+   readiness. `docs/dry-run-checklist.md` is a separate manual platform trial and must
+   not impose toy-project or Pincer-kit assumptions on this audit.
+2. Run `scripts/pincer-status.sh` to determine the selected PRD and which stages have run
+   (it reads `.prd/`, associated `tickets/`, `NOTES.md`; add `git log`). If `$ARGUMENTS`
+   names a stage, check only up
    to that stage.
 3. Check every applicable item mechanically where possible:
    - File existence and frontmatter: read the files.
    - Commit format and story: `git log --oneline`.
-   - Receipts: every done ticket carries `verified:` (a status warning means one was
-     marked done by hand). Re-run at least two checks with
-     `scripts/pincer-ticket.sh verify T-{NN}` — on a done ticket it re-checks without
-     touching the receipt — and report actual output.
+   - Receipts: every done ticket carries current `last_check` and `verified` evidence;
+     any status warning fails the audit. Do not call `pincer-ticket.sh` from Release:
+     it writes receipts and would invalidate the evaluated candidate.
+   - Run the repository's candidate-wide release gate directly (`npm test`, or the
+     equivalent declared by the project) and report its actual output. Any failure
+     blocks PASS. Confirm `git status --short` remains clean afterward.
 4. For judgment items (tickets genuinely S/M, history reads as a story), give your
    verdict AND one sentence of evidence — never a bare pass.
 5. Present a table: checklist item | pass/fail/skipped | evidence. Order by stage.

package/template/.claude/commands/pincer-status.md

@@ -12,7 +12,7 @@ start of a session. Read-only: change nothing.
 
 1. Run `scripts/pincer-status.sh`. It reads the artifacts on disk (`.prd/`, `tickets/`,
    `NOTES.md`) and prints the PRD state, every ticket with its state and clock-based
-   elapsed time, what is blocked, build time against the budget, any warnings (a ticket
+   elapsed time, what is blocked, build time against any explicit user budget, any warnings (a ticket
    marked done without a verification receipt), and the next command to run.
 2. Report in three lines: where the workflow is, what is in progress or blocked, and the
    next command. Quote the `Next` line as-is.

package/template/.claude/hooks/block-dangerous.sh

@@ -1,20 +1,9 @@
 #!/bin/bash
-# PreToolUse guardrail for Bash commands. Blocks destructive or pipe-to-shell
-# patterns regardless of permission mode — these are never auto-approved; a
-# human runs them in their own terminal if truly intended.
-# Exit 2 blocks the tool call; stderr goes back to the agent.
-
-input=$(cat)
-
-if printf '%s' "$input" | grep -qE \
-  -e 'rm -rf?( -[a-z]+)* +(/|~|\$HOME)' \
-  -e 'git push[^|;&]*(--force|-f)( |$)' \
-  -e '(curl|wget)[^|;&]*\|[[:space:]]*(sudo )?(ba|z)?sh' \
-  -e 'chmod( -R)? +777' \
-  -e '--dangerously-skip-permissions' \
-  -e 'git reset --hard[^|;&]*origin/'; then
-  echo "Blocked by PINCER guardrail: destructive or pipe-to-shell command. If this is genuinely intended, the user runs it manually in their own terminal." >&2
+# PreToolUse guardrail for Bash commands. The Node helper parses the hook JSON
+# before inspecting the actual command and recognizes the documented command
+# forms without matching harmless strings elsewhere in the payload.
+command -v node >/dev/null 2>&1 || {
+  echo 'Blocked by PINCER guardrail: Node.js is required to parse hook input safely.' >&2
   exit 2
-fi
-
-exit 0
+}
+exec node "$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)/hook-policy.cjs" dangerous

package/template/.claude/hooks/hook-policy.cjs

@@ -0,0 +1,258 @@
+#!/usr/bin/env node
+const fs = require('node:fs');
+const path = require('node:path');
+
+const mode = process.argv[2];
+const input = fs.readFileSync(0, 'utf8');
+
+function block(message) {
+  const label = mode === 'ticket' ? 'ticket guard' : 'guardrail';
+  process.stderr.write(`Blocked by PINCER ${label}: ${message}\n`);
+  process.exit(2);
+}
+
+let payload;
+try {
+  payload = JSON.parse(input);
+} catch {
+  block('hook input is not valid JSON.');
+}
+if (!payload || typeof payload !== 'object' || Array.isArray(payload) ||
+    typeof payload.tool_name !== 'string' || !payload.tool_input ||
+    typeof payload.tool_input !== 'object' || Array.isArray(payload.tool_input)) {
+  block('hook input must contain tool_name and a tool_input object.');
+}
+
+function lexShell(source) {
+  const tokens = [];
+  let word = '';
+  let quote = '';
+  const push = () => { if (word) tokens.push({ value: word }); word = ''; };
+  for (let i = 0; i < source.length; i++) {
+    const ch = source[i];
+    if (quote) {
+      if (ch === quote) quote = '';
+      else if (ch === '\\' && quote === '"' && i + 1 < source.length) word += source[++i];
+      else word += ch;
+      continue;
+    }
+    if (ch === "'" || ch === '"') { quote = ch; continue; }
+    if (ch === '\\' && i + 1 < source.length) { word += source[++i]; continue; }
+    if (/\s/.test(ch)) {
+      push();
+      if (ch === '\n') tokens.push({ op: ';' });
+      continue;
+    }
+    if ('|;&><()'.includes(ch)) {
+      push();
+      const pair = source.slice(i, i + 2);
+      if (['||', '&&', '>>', '<<'].includes(pair)) { tokens.push({ op: pair }); i++; }
+      else tokens.push({ op: ch });
+      continue;
+    }
+    word += ch;
+  }
+  push();
+  return tokens;
+}
+
+function shellCommands(source) {
+  const commands = [];
+  let words = [];
+  let operators = [];
+  const finish = (separator = '') => {
+    if (words.length || operators.length) commands.push({ words, operators, separator });
+    words = []; operators = [];
+  };
+  for (const token of lexShell(source)) {
+    if (token.value !== undefined) words.push(token.value);
+    else if (['|', ';', '&&', '||', '(', ')'].includes(token.op)) finish(token.op);
+    else operators.push(token.op);
+  }
+  finish();
+  return commands;
+}
+
+function commandParts(command) {
+  const words = [...command.words];
+  let i = 0;
+  while (i < words.length && /^[A-Za-z_][A-Za-z0-9_]*=/.test(words[i])) i++;
+  if (words[i] === 'env') {
+    i++;
+    while (i < words.length && (/^[A-Za-z_][A-Za-z0-9_]*=/.test(words[i]) || words[i].startsWith('-'))) i++;
+  }
+  if (words[i] === 'sudo') {
+    i++;
+    while (i < words.length && words[i].startsWith('-')) i++;
+  }
+  while (['command', 'builtin', 'exec'].includes(words[i])) i++;
+  const executable = words[i] ? path.basename(words[i]) : '';
+  return { executable, args: words.slice(i + 1), words };
+}
+
+function gitSubcommand(args) {
+  let i = 0;
+  while (i < args.length) {
+    if (['-C', '-c', '--git-dir', '--work-tree'].includes(args[i])) i += 2;
+    else if (args[i].startsWith('-')) i++;
+    else return { name: args[i], args: args.slice(i + 1) };
+  }
+  return { name: '', args: [] };
+}
+
+function dangerousReason(source, depth = 0) {
+  if (depth > 2) return '';
+  const commands = shellCommands(source);
+  for (let i = 0; i < commands.length; i++) {
+    const { executable, args, words } = commandParts(commands[i]);
+    if ((executable === 'claude' || (executable === 'npx' && args.some(arg => arg === 'claude'))) &&
+        words.includes('--dangerously-skip-permissions')) return 'permission bypass flags are not allowed.';
+    if (executable === 'git') {
+      const sub = gitSubcommand(args);
+      if (sub.name === 'push' && sub.args.some(arg => arg === '-f' || arg === '--force' || arg.startsWith('--force=')))
+        return 'force-push commands are not allowed.';
+      if (sub.name === 'reset' && sub.args.includes('--hard') && sub.args.some(arg => /^origin\//.test(arg)))
+        return 'hard resets to a remote branch are not allowed.';
+    }
+    if (executable === 'rm') {
+      let recursive = false, force = false;
+      const targets = [];
+      for (const arg of args) {
+        if (arg === '--recursive') recursive = true;
+        else if (arg === '--force') force = true;
+        else if (/^-[^-]/.test(arg)) { recursive ||= /[rR]/.test(arg.slice(1)); force ||= /f/.test(arg.slice(1)); }
+        else if (arg !== '--') targets.push(arg);
+      }
+      if (recursive && force && targets.some(arg => arg.startsWith('/') || arg === '~' || arg.startsWith('~/') || arg === '$HOME' || arg.startsWith('$HOME/') || arg.startsWith('${HOME}')))
+        return 'recursive forced deletion of an absolute or home path is not allowed.';
+    }
+    if (executable === 'chmod' && args.some(arg => /^(0?777|a\+rwx)$/.test(arg)))
+      return 'mass permission changes are not allowed.';
+    if (['sh', 'bash', 'zsh'].includes(executable)) {
+      const c = args.indexOf('-c');
+      if (c !== -1 && typeof args[c + 1] === 'string') {
+        const nested = dangerousReason(args[c + 1], depth + 1);
+        if (nested) return nested;
+      }
+    }
+    if (['curl', 'wget'].includes(executable) && commands[i].separator === '|') {
+      const next = commandParts(commands[i + 1] || { words: [] }).executable;
+      if (['sh', 'bash', 'zsh'].includes(next)) return 'downloading content directly into a shell is not allowed.';
+    }
+  }
+  return '';
+}
+
+const TICKET_PATH = /(^|[\\/])tickets[\\/]T-[0-9]+[^\\/]*\.md$/;
+const PROTECTED = ['status', 'started', 'last_check', 'verified', 'finished'];
+
+function ticketPath(value) {
+  return typeof value === 'string' && TICKET_PATH.test(value);
+}
+
+function stateFields(content) {
+  const result = Object.fromEntries(PROTECTED.map(key => [key, []]));
+  let scope = String(content ?? '');
+  if (scope.startsWith('---\n')) {
+    const end = scope.indexOf('\n---', 4);
+    if (end !== -1) scope = scope.slice(4, end + 1);
+  }
+  for (const line of scope.split(/\r?\n/)) {
+    const match = line.match(/^([a-z_]+):[ \t]*(.*)$/);
+    if (match && PROTECTED.includes(match[1])) result[match[1]].push(match[2]);
+  }
+  return result;
+}
+
+const sameState = (a, b) => JSON.stringify(stateFields(a)) === JSON.stringify(stateFields(b));
+
+function absoluteToolPath(file) {
+  const root = process.env.CLAUDE_PROJECT_DIR || process.cwd();
+  return path.isAbsolute(file) ? file : path.resolve(root, file);
+}
+
+function existingContent(file) {
+  try { return fs.readFileSync(absoluteToolPath(file), 'utf8'); }
+  catch { return null; }
+}
+
+function applyEdit(content, oldText, newText, replaceAll = false) {
+  if (typeof oldText !== 'string' || typeof newText !== 'string') block('edit payload must contain string old_string and new_string values.');
+  if (!oldText) block('edit old_string must not be empty.');
+  if (replaceAll) return content.split(oldText).join(newText);
+  const index = content.indexOf(oldText);
+  return index === -1 ? content : content.slice(0, index) + newText + content.slice(index + oldText.length);
+}
+
+function guardEdits(tool, toolInput) {
+  const file = toolInput.file_path;
+  if (typeof file !== 'string') block(`${tool} payload must contain a string file_path.`);
+  if (!ticketPath(file)) return;
+  const before = existingContent(file);
+  if (tool === 'Write') {
+    if (typeof toolInput.content !== 'string') block('Write payload must contain string content.');
+    if (before !== null) {
+      if (!sameState(before, toolInput.content)) block('ticket lifecycle fields may only be changed by pincer-ticket.sh.');
+    } else {
+      const state = stateFields(toolInput.content);
+      if (state.status.length !== 1 || state.status[0].split(/[ \t]+#/)[0].trim() !== 'open' ||
+          PROTECTED.slice(1).some(key => state[key].length))
+        block('new tickets must start with status: open and no lifecycle receipts or timestamps.');
+    }
+    return;
+  }
+  const edits = tool === 'MultiEdit' ? toolInput.edits : [toolInput];
+  if (!Array.isArray(edits)) block('MultiEdit payload must contain an edits array.');
+  let after = before ?? '';
+  for (const edit of edits) {
+    if (!edit || typeof edit !== 'object') block('each edit must be an object.');
+    after = applyEdit(after, edit.old_string, edit.new_string, edit.replace_all === true);
+  }
+  if (!sameState(before ?? '', after)) block('ticket lifecycle fields may only be changed by pincer-ticket.sh.');
+}
+
+function isExactPincerCall(source) {
+  const commands = shellCommands(source).filter(command => command.words.length);
+  if (commands.length !== 1 || commands[0].separator) return false;
+  const { executable, args } = commandParts(commands[0]);
+  let words = [executable, ...args];
+  if (['bash', 'sh'].includes(words[0])) words = words.slice(1);
+  if (!/pincer-ticket\.sh$/.test(words[0] || '')) return false;
+  const action = words[1];
+  if (!['start', 'verify', 'done', 'bind'].includes(action)) return false;
+  return action === 'bind' ? words.length === 4 : words.length === 3;
+}
+
+function ticketShellMutation(source) {
+  if (isExactPincerCall(source)) return false;
+  const commands = shellCommands(source);
+  for (const command of commands) {
+    const { executable, args, words } = commandParts(command);
+    const hasTicket = words.some(ticketPath) || /(^|[\s'"`])tickets[\\/]T-[0-9]+[^\s'"`]*/.test(source);
+    if (!hasTicket) continue;
+    if (command.operators.some(op => op === '>' || op === '>>')) return true;
+    if (['rm', 'mv', 'cp', 'install', 'truncate', 'touch', 'tee', 'ed', 'ex'].includes(executable)) return true;
+    if (['python', 'python3', 'node', 'ruby'].includes(executable)) return true;
+    if (['sed', 'perl'].includes(executable) && args.some(arg => /^-[^-]*i/.test(arg) || arg === '--in-place' || arg.startsWith('--in-place='))) return true;
+    if (executable === 'git') {
+      const sub = gitSubcommand(args).name;
+      if (['checkout', 'restore', 'reset', 'clean'].includes(sub)) return true;
+    }
+  }
+  return false;
+}
+
+if (mode === 'dangerous') {
+  if (payload.tool_name !== 'Bash') process.exit(0);
+  if (typeof payload.tool_input.command !== 'string') block('Bash payload must contain a string command.');
+  const reason = dangerousReason(payload.tool_input.command);
+  if (reason) block(reason);
+} else if (mode === 'ticket') {
+  if (['Edit', 'Write', 'MultiEdit'].includes(payload.tool_name)) guardEdits(payload.tool_name, payload.tool_input);
+  else if (payload.tool_name === 'Bash') {
+    if (typeof payload.tool_input.command !== 'string') block('Bash payload must contain a string command.');
+    if (ticketShellMutation(payload.tool_input.command)) block('shell commands may not write or reset ticket files; use pincer-ticket.sh for lifecycle state.');
+  }
+} else {
+  block('hook policy mode is invalid.');
+}